What Is a Word?

What Is a Word?
Since the beginnings of grammatical study in Europe, the concept of a word has been
considered to be of central importance. But there are several ways of defining words, and
these are not equivalent. We therefore need to examine the several definitions, and to
understand the differences among them. Otherwise, we will have no chance of providing
sensible answers to such questions as the following: How many words are there in
English? Are English dog and dogs the same word or different words?
It is usual to distinguish four definitions of ‘word’. We will look at each of these
first from the point of view of English. Then we will examine some data from other
languages. Finally, we will look at some special cases.
1. Orthographic words. An orthographic word is a written sequence which has a
white space at each end but no white space in the middle.
If you look at the paragraph above, you will find it easy to pick out the
orthographic words: Since, the, beginnings, of, and so on. Very obviously, orthographic
words exist only in written texts, and they have no existence in speech.
Consider the sequence ice cream. As we will see later, this is in some important
respects a single English word, but nevertheless it is two orthographic words, not one. If
we wanted to, we could write it as ice-cream, in which case it would be one orthographic
word, but this spelling is not usual.
Such fluctuation is by no means rare in English, however. We might choose to
write any of land owners, land-owners or landowners. All three forms have the same
meaning, but the first is two orthographic words, while the second and third are one
orthographic word each. We encounter both pocket knife and pocketknife, common sense
and commonsense, toy shop and toyshop, and hundreds of others of the same kind. The
rules of English orthography simply do not specify which compounds should be written
with a white space and which not, and individual preferences vary considerably.
Of course, this freedom to choose where to put white spaces is far from absolute.
In most cases, the rules of our orthography dictate where the white spaces should go, and
failure to conform produces manifest illiteracy.
However, orthographic words are generally of very little linguistic interest. They
are important in learning to write a language in an educated manner, and they may be of
interest in a linguistic study of writing systems. But, for most linguistic purposes,
orthographic words are irrelevant.
What about other languages? The concept of orthographic words can be readily
extended to any writing system which uses white spaces more or less as we use them in
English. But not all writing systems do this. Most ancient languages, including
Phoenician, Greek and Latin, were very often written with no white spaces or other
dividers. Here is an example from Greek.
TOMNHMATODEP
OIHSATOSIDARIOSPARMENO
NTOSUIOSEAUTWIKAIGUNA
IKIKAIUIWIPUBIALHI
If this passage were divided into words in the way that later became conventional in
Greek, it would look like this:
TO MNHMA TOD EPOIHSATO SIDARIOS PARMENONTOS UIOS EAUTWI KAI G
UNA IKI KAI UIWI PUBIALHI
You can see that the concept of orthographic words is of no use with such a writing
system. The idea of separating words was slow to take hold among the ancients, but
eventually some peoples began to write their languages with dots or strokes to separate
the words. Here is an example from Oscan, an ancient language of Italy (the original text
was written right-to-left, but we’ll ignore that here):
STATUS•PUS•SET•HURTIN•KERRIIIN•VEZKEI•STATIF•...
Here the dots are doing the same job as the white spaces in English, and so we can
usefully talk about orthographic words in this text.
2. Phonological words. A phonological word is a piece of speech which behaves as a
unit of pronunciation according to criteria which vary from language to language.
In English, the most useful criterion is this one: a phonological word contains
only one main stress. Consider the following sentence, as you would pronounce it in
relaxed colloquial speech:
The rest of the books’ll have to go here.
There are five main stresses here, falling on the words rest, books, have, go and here.
This sentence therefore contains five phonological words. One obvious way of breaking
up the utterance into phonological words is as follows:
[the rest] [of the books’ll] [have to] [go] [here]
(You might argue for [‘ll have] [to go], but we won’t pursue that issue here.) You can
see that not all of these units correspond to units that we might want to recognize for
other purposes: for example, of the books’ll is certainly not a unit of grammatical
structure, but we nevertheless pronounce it as a single phonological word.
In other languages, phonological words may be identified by other criteria.
Turkish, for example, has a phenomenon called vowel harmony, by which a single
phonological word is permitted to contain only certain sequences of vowels. As a
consequence, each Turkish suffix exhibits multiple forms differing in vowel quality. For
example, the suffix meaning ‘in’ appears as either -de or -da, according to the
requirements of harmony: ev ‘house’ forms evde ‘in the house’, but oda ‘room’ forms
odada ‘in the room’. And the suffix meaning ‘my’ has four forms: ev ‘house’ gives evim
‘my house’; at ‘horse’ gives atIm ‘my horse’; göz ‘eye’ gives gözüm ‘my eye’; and top
‘gun’ gives topum ‘my gun’.
Phonological words are important in the study of pronunciation, but they are
irrelevant to the study of grammar.
3. Lexical items. A lexical item (or lexeme) is an abstract unit of the lexicon
(vocabulary) of a language, with a more or less readily identifiable meaning or function.
A lexical item is a word in the sense in which a dictionary contains words. It is
also the sense of ‘word’ in cases like “How many words are present in the vocabulary of
an ordinary English-speaker?” and “I learned twenty new words of French today.”
A lexical item is an abstract unit, and it must be represented in speech or writing
by one of the possibly several forms it can assume for grammatical purposes. For
example, if we want to mention canine animals, we must use either the singular form dog
or the plural form dogs. But these two grammatical forms both represent the same single
abstract unit, the same lexical item. We can conveniently represent that lexical item as
DOG. Then dog and dogs are the two possible forms of the lexical item DOG.
In the same way, we recognize a lexical item TAKE, which can be represented by
any of the five grammatical forms take, takes, took, taken and taking.
A dictionary provides entries for lexical items. So, for example, we do not expect
to find separate entries in the dictionary for dog and dogs: we expect to find only one
entry for the lexical item DOG. Likewise, we expect to find only one entry for the lexical
item TAKE, and not five entries for the five forms of that lexical item.
(A small complication. Some dictionaries have a policy of providing entries for
irregular forms like took, but an entry for one of these is purely a cross-reference to the
main entry. So, the entry for took will say merely “see take”.)
In English, the lexical item BE, uniquely, has eight different forms, but no other
English lexical item has more than five, and most have fewer than this – sometimes only
one. However, in grammatically more elaborate languages, such as French, Russian,
Latin and Arabic, a lexical item may have several dozen different forms. And there exist
languages in which certain lexical items (usually verbs) can have several hundred forms.
3.1. Citation forms. We need to talk about lexical items far more often than about other
kinds of words. But we have a problem. A lexical item is an abstract unit, and it must
always appear in one or another of its various forms. Therefore, we are obliged to choose
one of those forms to represent the lexical item when we want to talk about it. The form
we choose is called the “citation form” of the lexical item. The citation form (or
dictionary form) of a lexical item is the particular grammatical form of it which we use
in naming it, talking about it, and entering it in a dictionary.
In English, we are exceptionally fortunate. Almost every English lexical item has
one form which carries no grammatical marking at all. This form, the base form, is the
natural choice for the citation form.
For a noun like DOG, we choose the singular dog as the citation form, and not the
plural dogs. For the verb TAKE, we choose the infinitive take, and not an inflected form
like taking or took. For the adjective BIG, we choose the positive form big, and not an
inflected form like bigger. For the preposition WITH, there is nothing to talk about: this
lexical item has only the single form with, and this is therefore its citation form.
A further problem may arise with lexical items which are defective. A defective
lexical item is one that lacks some of the forms normally exhibited by a lexical item of its
class. For example, the nouns FURNITURE and HAPPINESS are defective, since they
have no plural forms, while the nouns OATS and POLICE have no singular forms. And
the verbs MUST and BEWARE both lack a number of the forms exhibited by most verbs.
If one of the missing forms happens to be the one we would ordinarily choose as the
citation form, then we are forced to make a different choice. For example, since OATS
has no singular, we must choose the plural form oats as the citation form.
In other languages, we are not always so fortunate as to find one form of a lexical
item which carries no grammatical marking at all. For example, in the Romance
languages, such as French, Spanish and Italian, a verb has no form which carries no
marking at all: every verb form in these languages bears a grammatical marking. We are
therefore obliged to choose as the citation form one of these grammatically-marked
forms. In all these languages, the conventional choice is the infinitive, which, unlike the
English infinitive, bears a grammatical ending. So, for example, the verb meaning ‘sing’
is cited as French chanter, Spanish cantar and Italian cantare.
The position is much the same in Latin. We might therefore choose the infinitive
as the citation form of a verb, and linguists occasionally do this, but, among classicists,
the conventional citation form is the first-person singular present indicative active form.
So, the verb for ‘sing’ is cited, not as its infinitive canere, but as cano ‘I sing’, and this is
the form under which the verb is entered in a Latin dictionary.
Much the same occurs in Arabic, but an Arabic verb has no infinitive. The
conventional citation form of an Arabic verb is its masculine third-person singular pasttense form. So, for example, the verb meaning ‘write’ is cited as kataba ‘he wrote’, and
this is the form under which the verb is entered in a dictionary.
4. Grammatical word-forms. A grammatical word-form (or GWF, or grammatical
form) is one of the several forms that may be assumed by a lexical item for grammatical
purposes.
So, for example, dog and dogs are grammatical forms of the lexical item DOG,
and take, takes, took, taken and taking are all grammatical forms of the lexical item
TAKE. Sometimes a lexical item exhibits only one grammatical form. For example, the
lexical items OATS, POLICE, BEWARE and WITH have only one grammatical form
apiece: oats, police, beware and with. Nevertheless, it may still be useful to distinguish
the lexical item OATS from its sole grammatical form oats, and likewise for the others.
We have already seen that an English lexical item usually has one grammatical
form which carries no grammatical marking at all, its base form. All the other
grammatical forms of a lexical item, the ones that carry grammatical markings, are the
inflected forms of that lexical item. So, for example, DOG has the base form dog and
the inflected form dogs, while TAKE has the base form take and the inflected forms
takes, took, taken and taking. But POLICE, BEWARE and WITH have no inflected
forms at all, while OATS has only the inflected form oats. (A complication: since the
form police is plural, it must bear an abstract inflection for plurality. Some linguists
might therefore prefer to say that police is an inflected form, even though it carries no
overt marking.)
In some analyses, the number of grammatical forms exhibited by a lexical item
may be larger than the number of overtly distinct forms. Here is a simple example.
Some English verbs exhibit different forms for the past tense and the past participle. The
verb TAKE is one of these: She took the exam, but She has taken the exam. Many verbs,
however, present identical forms forms for the past tense and the past participle. An
example is FINISH: She finished the exam, but She has finished the exam. Since the
distinction between the past tense and the past participle is essential for English verbs in
general, most linguists prefer to say that past-tense finished and past-participle finished
are two different grammatical forms of the verb FINISH.
Here is a slightly less obvious example. For any English verb, the present-tense
forms (usually apart from the third-singular form) are identical to the infinitive. We can
see this with TAKE: You should take an umbrella (infinitive take), but I always take an
umbrella (present-tense take). Even though there is no English verb which exhibits
different forms for the infinitive and the present tense, many linguists (not all) would still
prefer to say that we are looking at two different grammatical forms here, because of the
difference in grammatical behaviour.
We finish with a third example, even more subtle. The -ing form of an English
verb has at least three different grammatical functions. We will illustrate these with
taking. First, taking can be a participle: Taking her umbrella, she slipped out of the
house. Second, it can be part of a progressive (continuous) verb-form: Susie is taking her
umbrella. Third, it can be a gerund: Taking an umbrella would be a good idea. Some
linguists (again, not all) would prefer to say that all three of these functions represent
different grammatical forms of TAKE, even though the three forms are always identical
for every verb.
These decisions are not matters of absolute truth: they are merely matters of
analytical taste and convenience, and linguists vary in their preferences.
By the way, there is one more item of the form taking. This is the verbal noun,
and it appears in the film title The Taking of Pelham 1-2-3. But this taking is not a
grammatical form of the verb TAKE, and in fact it’s not a verb-form at all: it’s a noun,
and it represents a distinct lexical item TAKING. This noun TAKING is related to the
verb TAKE by a derivational process, much as the noun ARRIVAL is related to the verb
ARRIVE. See the next section.
[A warning. Quite a few writers, including some who should know better,
confuse verbal nouns in -ing with gerunds, and apply the label “gerunds” to all of them
indiscriminately. This is not a practice you should imitate. The gerund taking is a verbform (for example, it can take an object, as in the example above), and it is an inflected
grammatical form of the verb TAKE. However, the verbal noun taking is not a verbform at all, but a noun: it behaves like a noun, and it has no verbal properties at all.]
5. Inflection and derivation. We have seen that a lexical item can appear in several
grammatical forms, some of which carry inflections (overt grammatical markings). It is
essential to distinguish the process of inflection from the quite different process called
derivation.
Inflection is the variation in form of a lexical item for grammatical purposes.
Derivation is the construction of a new lexical item from another lexical item,
usually by the addition of an affix (a prefix or a suffix).
Take the lexical item CAT, which has two grammatical forms: cat and cats. But
what about the word catty, as in a catty remark? How should we classify this?
This is not another inflected form of CAT. Instead, it is a different lexical item,
CATTY. How do we know this? For one thing, CAT is a noun, and all its grammatical
forms are nouns, while CATTY is an adjective.
The relation between CAT and CATTY is one of derivation: we say that CATTY
is derived from CAT by the addition of a suffix.
Since CATTY is a different lexical item from CAT, we expect to find CATTY
entered in the dictionary, and our expectations are fulfilled: look at any dictionary of
English.
In the same way, DOGLIKE is a different lexical item from DOG, and it too
deserves its own entry in the dictionary. A few more examples of derivation by
suffixation are HAPPINESS from HAPPY, ARRIVAL from ARRIVE, SINGER from
SING, SLOWLY from SLOW, KINGDOM from KING, MOTHERHOOD from
MOTHER, NUDISM from NUDE and REDDISH from RED. Some examples involving
prefixation are REWRITE from WRITE, UNHAPPY from HAPPY, NON-SMOKER
from SMOKER and PRE-WAR from WAR.
Some examples of derivation are irregular and messy, such as DESTRUCTION
from DESTROY, CALCULABLE from CALCULATE, COMPREHENSIBLE from
COMPREHEND and MEDICAL from MEDICINE.
In principle, we expect to find a separate entry for every lexical item obtained by
derivation. In practice, however, dictionaries often cut corners. When a derived lexical
item has a regular form and a transparent meaning, as with UNHAPPY, HAPPINESS,
REWRITE and SLOWLY, a dictionary may elect to save space by not providing an
ordinary entry for the word. A large dictionary may simply list the word without
definition, either at the end of the entry for its source word (look for slowly under slow),
or at the bottom of the page (look at the words in un- in a big dictionary), while a small
dictionary may omit the word altogether. This is done merely to save money, and not to
deny that all these words are genuine lexical items which really deserve their own entries.
English exhibits an interesting type of derivation called zero-derivation, or
conversion, by which a lexical item is simply shifted from one word-class to another,
without adding any material. For example, the adjective brown, as in brown shoes, has
been converted into a verb, as in brown the meat, and the noun shoe, as in new shoes, has
been converted into a verb, as in shoe the horses. English allows this with some freedom,
and examples can be elaborate. For example, the noun smoke, as in full of smoke, has
been converted into a verb, as in Susie smokes, and this verb has been converted into a
different noun, as in Susie was having a smoke.
In all these cases, we must regard the zero-derived word as belonging to a
different lexical item from the source, because it belongs to a different word-class. So,
we must recognize a lexical item BROWN1, which is an adjective, and another lexical
item BROWN2, which is a verb, and similarly for the other cases.
In the examples we have just been looking at, it is obvious which items have been
derived from which other items. But the direction of derivation is not always obvious.
Take kiss. We have a verb KISS, as in Susie kissed Mike, and a noun KISS, as in Susie
gave Mike a kiss. Presumably one of these is derived from the other. But which one is
it? This is not obvious, and we may have to make an arbitrary decision.
6. Multi-part and discontinuous words. English and some other languages exhibit
certain items which present awkward problems from the point of view of wordhood.
These are the items which appear to be words by some criteria, but which exist in two or
three pieces, sometimes separated by other material.
The most obvious English examples are the phrasal verbs. A phrasal verb
consists of a simple verb plus a particle, or rarely two particles. Examples are make up,
take off, turn on, ring up and put up with. A phrasal verb represents two (or three)
orthographic and phonological words, but it must usually be considered a single lexical
item, since its meaning is often unpredictable from the meanings of its components.
Consider the several items of the form make up: She made up a story; She made up her
face; She made up with her boyfriend; She made up the numbers.
Many phrasal verbs can occur discontinuously: She took off her dress or She took
her dress off; She turned on the light or She turned the light on; and so on.
It appears, then, that we must recognize lexical items which occur in pieces,
sometimes in pieces which are separated by other material.
English has another class of awkward verbs, the prepositional verbs. A
prepositional verb consists of a simple verb (possibly plus an adverb) plus a preposition.
Examples are call on, look at, take to, and look down on: We called on Susie; Look at the
rainbow!; I didn’t take to our new neighbours; Susie looks down on New Agers.
Prepositional verbs differ from phrasal verbs in ways that are explained in reference
grammars of English.
In these examples, sequences like on Susie and on New Agers appear to be
syntactic units, prepositional phrases. Nevertheless, it appears that we must recognize
call on (‘visit’) and look down on (‘regard with contempt’) as single lexical items,
because of their meanings. We are thus faced with conflicting analyses: the item on is
syntactically part of a prepositional phrase, but lexically part of a multi-part verb.
Examples like these present formidable problems of analysis, and they stretch our
conception of words to the breaking point.
Other languages present further examples of multi-part and discontinuous words.
Consider Mandarin Chinese. In Chinese, ‘(the) can’ is guòn, but ‘into the can’ is dào
guòn li‡, where ‘into’ is expressed by the discontinuous item dào...li‡.
7. Content words and grammatical words. Lexical items are commonly divided unto
content words and grammatical words.
A content word (or full word, in the Chinese terminology) is a lexical item
which has semantic content – that is, which has a readily identifiable meaning.
A grammatical word (or empty word) has little or no identifiable meaning, but
has one or more grammatical functions.
A content word can be defined, and it can often be translated into another
language with some ease. English content words include house, dog, give, big and
carefully. An English dictionary will provide reasonable definitions of these items. And
we can say that house is equivalent to Welsh ty^, French maison, Spanish casa, Turkish
ev, Swahili nyumba and Japanese ie. (But note that, because of lexical differences
among languages, not all content words can be translated quite so straightforwardly.)
A grammatical word cannot be defined, and looking for an equivalent in another
language is often pointless. English grammatical words include of, the, and, have, who
and if. A dictionary cannot provide a definition of one of these, but can only give an
account of its grammatical functions: asking for the “meaning” of the word of is a waste
of time. The English phrase a bottle of wine contains the two grammatical words a and
of; its Basque translation boteila bat ardo is literally ‘bottle a wine’, with no equivalent of
of; and its Welsh equivalent potel gwin is literally ‘bottle wine’, with no equivalent of a
or of.
Some words appear to be borderline in this classification, with clear grammatical
functions but some degree of identifiable semantic content, such as in, with, we, this and
for. The best policy is to classify these items as grammatical words, since their
grammatical functions are usually much more important than any traces of meaning they
may possess.
8. Clitics. A clitic is an item which represents a lexical item and a grammatical wordform but which does not make a phonological word by itself.
Since a clitic cannot form a phonological word by itself, it must be phonologically
bound to a more substantial item, its host, with which it forms a phonological word,
possibly together with other clitics.
Among the English clitics are the articles a(n) and the, which are bound to a
following host, as in a book and the man, and the auxiliary ‘ll, which is bound to a
preceding host, as in John’ll do it.
French possesses a large number of clitics. In the French sentence Il te le donnera
‘He’ll give it to you’, the pronouns il ‘he’, te ‘you’ and le ‘it’ are all clitics bound to the
following verb.
Turkish has a clitic de ~ da ‘also, too’, which is bound to a preceding item and
must agree with that item in vowel harmony: siz de ‘you too’, but onlar da ‘they too’.
A clitic which precedes its host, like the English articles and the French pronouns,
is a proclitic. A clitic which follows its host, like English ‘ll and the Turkish item, is an
enclitic.
Some linguists would suggest that clitics are items whose claim to wordhood is
somewhat marginal. From a historical point of view, a clitic is perhaps to be seen as an
item which has lost its status as an autonomous word, but which has not yet been reduced
to the status of an affix (a prefix or a suffix).
9. Short forms. English exhibits a number of short forms of several kinds, and these
short forms merit some discussion from the point of view of wordhood. Among nonlinguists, there is a habit of labelling all or most of these short forms “abbreviations”, but
this practice is misconceived. We will distinguish true abbreviations from the other
cases.
9.1. Abbreviations. An abbreviation is a short way of writing a word or a phrase, using
only letters of the alphabet and possibly full stops. An abbreviation is strictly a written
form. It has no pronunciation of its own, and it can only be pronounced by pronouncing
the full form which it abbreviates – or, in some cases, by spelling it out letter by letter.
Familiar English abbreviations include Prof. for Professor, Sgt for Sergeant, Dr
for Doctor (only as titles accompanying names), BC for before Christ, mph for miles per
hour, kg for kilograms and C for degrees Celsius. Very many of our common
abbreviations are derived from Latin: a.m. for in the morning, lb. for pounds, e.g. for for
example, i.e. for in other words, and etc. for and other things. The exceptional forms Mrs
and Ms are treated as abbreviations, even though no longer forms exist.
An abbreviation consisting of only one or two letters may sometimes be spelled
out letter by letter. This is commonplace with a.m. (“ay-em”) and BC (“bee-see”). In
other cases, this practice is usual only in reading out a written text, as with e.g. (“ee-jee”)
and i.e. (“eye ee”), and saying “ee-jee” or “eye-ee” in speech may be perceived as
eccentric.
An abbreviation usually qualifies as an orthographic word, but it is not a word in
any other sense. With a few marginal and idiosyncratic exceptions, an abbreviation is not
part of the spoken language at all.
9.2. Logograms. A logogram is a written character which is not a letter of the alphabet
but which conventionally represents a word, or rarely a sequence of words. Familiar
logograms include the digits, like ‘5’ for five, monetary symbols like ‘£’ for pounds, and
arithmetical symbols like ‘+’ for plus, ‘=’ for equals and ‘%’ for per cent. Another is
‘§’ for love, as in ‘I § NY’, but this one is confined to certain informal contexts.
A logogram is a representation of a lexical item or of a grammatical word-form.
It fits the definition of an orthographic word given above, but some linguists might prefer
to restrict that definition to cases consisting of letters of the alphabet, in which case a
logogram would not be an orthographic word.
9.3. Contractions. A contraction is a conventional brief way of pronouncing a sequence
of two (or rarely three) words which often occur together.
A contraction always has a distinct written form. Typical contractions are I’m for
I am, it’s for it is or it has, she’ll for she will, couldn’t for could not, won’t for will not,
hadn’t for had not, and (more informally) she’d’ve for she would have.
As you can see, some contractions have unexpected forms, such as won’t for will
not. And some contractions which were once common have become obsolete or regional,
such as ’tisn’t for it is not.
A contraction is always a single orthographic word and a single phonological
word. But it represents two (or three) lexical items, and two (or three) grammatical
word-forms.
9.4. Acronyms and initialisms. In contemporary English, we have become very fond of
coining new words in the following way: we begin with a phrase consisting of several
words; we extract the initial letters of the most important words in that phrase; and we put
the resulting sequence of letters together to form our new word – which almost always
has the same meaning as the original phrase.
Here are a few examples:
BBC
FBI
British Broadcasting Corporation
Federal Bureau of Investigation
NATO or Nato
AIDS or Aids
North Atlantic Treaty Organization
acquired immuno-deficiency syndrome
We may distinguish two kinds of outcome here, and there exist national differences in
terminology. Sometimes the resulting form can only be pronounced by spelling it out
letter by letter, as with BBC and FBI. In the British tradition, a formation of this kind is
an initialism. In other cases the resulting form can be pronounced like an ordinary word,
as with NATO and AIDS. In the British tradition, a formation of this kind is an acronym.
However, in the American tradition, the label “acronyms” is applied to all such
formations without distinction, and the term “initialism” is not used. British and
American books will therefore differ in their terminology.
There is a further trans-Atlantic difference. In all varieties of English, initialisms
like BBC and FBI are written entirely in capital letters. The Americans also write most
acronyms (in the British sense) entirely in capitals, and so they write NATO and AIDS.
But the British commonly write these things with only an initial capital, and the usual
British forms are Nato and Aids.
But not all acronyms are written with capitals. Some acronyms have become
perfectly ordinary lexical items, and they behave accordingly. For example, the phrases
self contained underwater breathing apparatus and light amplification by the stimulated
emission of radiation have given rise to the acronyms scuba and laser. Both of these are
ordinary English lexical items, entirely unremarkable apart from their origins.
The formation of laser illustrates the canonical pattern of formation, in which
small grammatical words are ignored in constructing the acronym. But many acronyms
are deliberately constructed in a non-canonical manner, in order to obtain a result which
can be easily pronounced. A good example is radar, from radio detection and ranging.
An initialism or an acronym is a lexical item. An acronym does not differ from
any other lexical item, except perhaps in its unusual written form. An initialism has both
an unusual written form and an unusual pronunciation, but otherwise it is an ordinary
lexical item.
Initialisms are sometimes confused with abbreviations, but they are not
abbreviations. First, an initialism always has its own pronunciation, distinct from the
pronunciation of the longer form which it represents. Abbreviations do not usually have
their own pronunciations. Second, an initialism, being a lexical item, can appear in a
structural position in a sentence in which a lexical item is appropriate. For example, we
can say or write the BBC’s decision, in which the initialism BBC bears the possesive
suffix -’s. No abbreviation can behave like this. We may occasionally encounter a
written – not spoken – form such as the sgt’s weapon, but this is not ordinary English
orthography: it is an example of shorthand, and shorthand is beyond the scope of our
discussion.
9.5. Clipped forms. A clipped form is an item which is obtained by extracting a piece
from a longer word or phrase. The process of extraction is clipping.
In all but very rare cases, a clipped form has the same meaning as the longer form
from which it is obtained. In contemporary English, we are inordinately fond of clipped
forms. We have gym from gymnasium, porn from pornography, flu from influenza,
fridge from refrigerator, phone from telephone, and gator from alligator. As these
examples show, any convenient part of a longer word may be clipped. Even
discontinuous pieces can be clipped, as with sci-fi for science fiction, sitcom for situation
comedy, biopic for biographical picture, and British maths for mathematics (compare
American math).
A clipped form is not an abbreviation. It is a genuine lexical item, just like any
other lexical item, and it is unusual in no way apart from its origin.
A clipped form accepts the grammatical inflections which are typical of its word
class. For example, nouns obtained by clipping can pluralize: gyms, fridges, phones,
gators. Verbs obtained by clipping can take ordinary verbal inflections. For example,
the noun disrespect has given rise to a noun diss, which in turn has yielded a verb diss
‘treat with disrespect’, and this verb behaves like any other verb: He was dissing me.
A clipped form can enter into compounds like any other lexical item: gym shoes,
porn star, fridge magnet, phone book. In some cases, the longer form would be abnormal
or impossible in the same position: who would say ??gymnasium shoes or ??pornography
star?
Clipped forms are entered in dictionaries like other lexical items, and they are
legal in Scrabble™.
Sometimes a clipped form displaces its original longer form. For example, the
clipped forms piano and bus have completely supplanted their sources, pianoforte and
omnibus, and mob has displaced its (Latin!) source mobile vulgus ‘the fickle crowd’.
And the words bra and cello are so close to replacing their source words brassière and
violoncello that many people do not even know the longer forms.
Note the following pair of examples. As observed above, English has an
abbreviation Prof. for Professor, as in the written form Prof. Chomsky. But we also have
a lexical item prof, obtained from professor by clipping, as in The Physics Department
plans to hire two new profs. The abbreviation Prof. and the lexical item prof are not the
same item at all, and they should not be confused.
10. Do all languages have words? Our intuitions suggest that a word is a unit which is
much smaller than a sentence, and that a sentence typically consists of a sequence of
words. But this account does not hold straightforwardly for all languages. In some
languages, it can be difficult to draw a distinction between sentences and words. Such
languages are known as polysynthetic languages.
Here is a typical sentence from Yup’ik, an Eskimo language of Alaska:
Kaipiallrulliniuk. ‘The two of them were apparently really hungry.’
The written form given here represents the ordinary pronunciation of this sentence. But,
at a somewhat more abstract level, we can analyses this sentence into a sequence of
structural units, or morphemes (if you haven’t yet learned about morphemes, you will be
meeting them soon):
kaig- ‘be hungry’
-piar- ‘really’
-llru- PAST
-llini- ‘apparently’
-u- INDICATIVE
-k ‘they two’
The sentence consists of a verb stem kaig- ‘be hungry’ followed by a string of suffixes.
In effect, then, the whole sentence is merely a grammatical form of this verb. As you
can see, the forms of some of these morphemes are altered when they are joined in
sequence.
In a sentence of this kind, there is nothing that we can call a word, except for the
sentence itself. And such sentences can be formidably long:
Ayaqaqucuaryuumiitqapiallruyugnarquq-qaa
‘I guess she probably didn’t really want to go for those short little trips, did she?’
We might wonder whether Yup’ik has any words at all, apart from its sentences. But not
all Yup’ik sentences are like these examples. Some sentences consist of shorter units
which we can reasonably call words:
Maurluqa ayunek pitlallruuq waten amllerivkenaki qillertaqluki enemun
agartaqluki.
‘My grandmother used to pick Labrador tea leaves, just a few like this, and tie
them
together and hang them inside the house.’
The words here may be glossed as follows:
maurluqa ‘my grandmother’
ayunek ‘Labrador tea leaves’
pitlallruuq ‘used to pick’
waten ‘like this’
amllerivkenaki ‘just a few’
qillertaqluki ‘and tie them together’
enemun ‘inside the house’
agartaqluki ‘and hang them’
So, Yup’ik does have words after all, though you can see that even the shortest Yup’ik
words appear to be more complex than most English words.